How Data Powers Michelin’s Journey Toward Sustainable Mobility

Michelin, a global leader in sustainable mobility, uses data and DevOps to drive innovation and efficiency. With insights from Yves Caseau, this interview explores Michelin's data-driven approach to growth, digital transformation, and customer-focused solutions.

Who Is Michelin?

Founded in 1889, Michelin began as a small rubber factory in France. Fast forward to today, the 132-year-old French company has become one of the world’s largest tire manufacturers and a leader in sustainable mobility. With nearly €650 million spent every year on R&D, Michelin produces tires for nearly every type of vehicle, such as bicycles, airplanes, farm equipment, heavy-duty trucks, motorcycles, as well as NASA’s space shuttle.

Michelin Expert Interview on Data in Digital Innovation

Yves Caseau, Group Chief Digital & Information Officer at Michelin, joins host Jedidiah Yueh (founder of Perforce Delphix) on the Data Company Podcast to share his perspective on the importance of software craftsmanship and the role data plays in digital innovation and business growth. Plus, he gives a preview of his latest book called “The Lean Approach to Digital Transformation.” Watch the full episode on YouTube.

Editor’s note: This interview was edited and condensed for brevity and clarity.

Behind the Book with Yves Caseau

Would you please share with us what inspired the book?

Yves Caseau: I joined Michelin three years ago as the Group CIO, and now I'm also in charge of digital. Before that, I was the CDO at AXA and even before that I worked at Bouygues Telecom. After 10 years, I wanted to share what I learned: what was easy and what was not so easy. The subtitle of the book is “From Customer to Code,” and the book says you have to master both. You have to be good at listening to your customers but also very good at producing software fast and efficiently using a DevOps approach. The book is about what I call exponential information systems.

It's how you manage your information system, so you can deliver exponential technologies to your company fast, frequently, constantly with a high level of quality that the digital players are very good at doing.

Traditional companies have to become software companies, and they have to become extremely good at this. Otherwise, they'll be in trouble.

One of the things you talk about in your book is that many enterprises struggle with digital transformation. Can you tell us a little bit more about that?

Yves Caseau: I'm a strong believer in execution and that you need digital talents. In the end, it’s about what you're engineering and your software. It starts with the business vision, but if you don't have the capabilities, you fail to deliver too slowly. By the time you're there, your market and users are not there anymore or there is someone who has disrupted you because he was much faster. The reason for writing this book was to first tell large companies that software craftsmanship matters and also that they have to be much more customer-centric than they are. The bigger the company, the more bureaucracy you have and long decision-making processes— which is the opposite of what you need to be customer centric.

Yves Caseau on Michelin & Data

Where do you think Michelin is along the journey of becoming a data company?

Yves Caseau: When I joined, I defined my higher strategy as the backbone of becoming a data-driven company with five layers. It's like a Maslow pyramid. You need to get better one step at a time. The first one was really quality of service, performance, and high availability. We've made progress. We reduced our downtime by a factor of five. The second layer was really simplification. I believe in moving software fast. It's like the law of physics. The more stuff you have, the slower you move. You need to reduce the weight and the complexity.

Here we've made progress. Then you need to have data that flows throughout the company. Michelin is a data-driven company, which has been very siloed. Today, it’s not only about having data but to make it available everywhere and fast. Then we need to move to system-to-system integration. Working on the edge with APIs was our purpose, and that was opening Michelin’s information system to the rest of the world. Here we're in progress. Lastly, quality of experience and user experience design. We're a product company, an engineering company, and a science company, which are strengths. But we are not so much of a user experience company, and we have to learn.

Jedidiah Yueh: It certainly sounds like customer experience is an area where data could be an advantage to Michelin. Are there other areas that you see where data plays a really important role for the future of the business?

Yves Caseau: Data plays a role everywhere, but I'm going to pinpoint two things. First, we are a product company, a chemical company. Our treasure is our IP about rubber, about metal. The way we have operated in the past, we extracted knowledge from a lot of tests. This is all about knowledge engineering and data engineering.

Our mission now is to do what we call data-driven engineering and data-driven R&D.

It's critical because the reason Michelin produces tires that have better performance than our competitors is because of our knowledge. If we don't continue to be good at knowledge engineering, we'll lose that. The other thing that is critical to us is because our tires are complex, the manufacturing is complex. Digital manufacturing is a wonderful opportunity to let digital handle some part of that complexity. Do I think we are where we should be? Not yet. But if I compare us to other companies in the manufacturing world or other tire manufactures, we are definitely quite good. From a strategy perspective, we have to be the best at chemical knowledge engineering with data.

Who owns data at Michelin?

Yves Caseau: Our businesses—like engineering, manufacturing, supply chain—own the data. That is quite nice because it means businesses have their own data strategy. However, it is too siloed. The challenge for Michelin is to move from everyone making their own data-driven decisions to everyone sharing data end-to-end across the product lifecycle, where the next gen of R&D leverages the data from usage, the next gen of manufacturing talks to R&D, and so on. We have to make our data models more interoperable because frankly, if I try to do an end-to-end throughout the lifecycle of the product matching with all the data that has been collected, it's possible but it's too much work today. Removing that data friction is really what it means to be data-driven.

Do you see bottlenecks in data management in your world and across other other customer environments?

Yves Caseau: I start with what matters the most for us, which is data quality. Data quality is very much related to the quality of service. When I looked at the root causes for why we had poor data, one third was about having the right tools to input data or process engineering. Two-thirds were about system failures, poor performance, desynchronization, batteries that run too late because you don't have the right version of the data and so on. We have made significant work to improve all of this, and our data quality has improved.

Then there is the obvious about data security which is critical to us because as I said we are like Coca-Cola, a chemistry company. Our values are recipes, and if we were to lose that, we would be in real trouble. Sometimes it makes us go a little slower, but that's an absolute must-do.

Our key data is more about tires than it is about sensitive data from customers. Our R&D knowledge is extracted from our testing data. If that data was stolen or extracted by a hacker, we would be in trouble, so we take that very seriously.

How strategically are you managing data today, and what is the road ahead?

Yves Caseau: We use a maturity model. It's inspired by acatech, which is the German Academy of Engineering. I'm a member of the French Academy of Engineering, so I like to look at those models. The first level is we use data to better see, understand, and predict what's going to happen. Once you master this, you use data to adapt and take the decision to automate. Once you become good at this, you start selling that capability to others. Through APIs, you are selling data as a product. At Michelin, we sell tires-as-a-service. We sell our data inside as a service, and that's the service where you start to make value from your data that you sell to others. The cost is by ton transported times kilometers. We are an active player in the connected tires industry, so I'm proud of what we do.

Yves Caseau on Speed, Downtime, & Transformation

Can you talk about lean software factories and the importance of APIs?

Yves Caseau: For us, APIs are key to our IS/IT strategy as a way to decouple, to trade capabilities, and to foster re-use. You have to move fast but you have to have capabilities, so you build capabilities through APIs and then you can move fast. The other thing which is really critical in a digital mindset is when you build an API, you build them for uses that you don't know. You have to think like a software company. Who is my community? What is my roadmap? What could people do with my data?

The lean software factory idea is really putting together agile and DevOps and producing software that changes constantly. It's not like 20 years ago when you said, “I'm going to write the spec, put in my code, do all the testing, and then I'm done.” This is not the way it works anymore. We're changing constantly, so you need a much higher level of excellence.

And what about your efforts in reducing down time?

Yves Caseau: The key lesson I learned from my previous years at Bouygues Telecom is that in the digital world, the quality of service is absolutely critical. It's a foundation. Many of our customers were saying, "Your tires are wonderful, but your systems are not as good as I expected." To get better, you have to reduce the mean time between failure and reduce the mean time to repair. The first one is a long-term journey. It's about better architecture, better systems, moving to the cloud, and leveraging the ways of the digital giant.

To improve the mean time to repair, it's about people, skills, and processes. We looked very hard at the Google SRE book because it's a good source of it. Then it's about tools. We are a happy Delphix customer because we have been using Delphix in production, either on our own to do hot fixes faster or to restore the database state to where it was before. Or when we work with a platform that somebody else is operating, we are able to capture the exact context, send it to our partners, and they can fix things faster—thanks to Delphix. I am happy to have this discussion because we have been very happy with using Delphix as a way to improve our quality of service.

How has Delphix contributed to digital transformation at Michelin?

Yves Caseau: We are making bets, and Delphix is one of our bets. I need scale and speed on how to make value from data. We selected Delphix because we like the technology as a tool to move data efficiently from one place to another, to reconstruct, to travel to the past. If we look at what we've done, we used Delphix to move data from our Exadata server legacy to community Linux servers to get more scalability and lower cost.

We are also using Delphix as a way to accelerate innovation and set up new test and development environments faster. We can get all the appropriate data very fast with a few pointers and a few clicks and have a virtual transfer of the data as opposed to a physical one.

Delphix Automates Data Delivery for Advanced Data-Driven Solutions

Legacy data delivery methods often rely on manual processes that consume valuable development time, creating roadblocks for organizations like Michelin in advancing innovation, transformation, and customer experience. Delphix automates these processes, delivering production-ready application data in just minutes.

Related blog >> What Is Delphix?

Data Virtualization Drives Michelin’s Data Strategy Forward

With Delphix, enterprises like Michelin break down data silos by syncing with production data sources and provisioning space-efficient virtual data copies for development, testing, or analytics. This approach accelerates provisioning times by 100x while reducing storage costs by 10x, paving the way for seamless data-driven engineering and collaboration.

Unlock Michelin-Level Data Efficiency Today

Virtual data copies from Delphix are not just space-efficient but are refreshed and versioned on demand, empowering DevOps teams to innovate. Combined with comprehensive data APIs, Delphix synchronizes data across environments for integration testing or rapid iterative improvements.

Accelerate Your Data Strategy Like Michelin

Request a Delphix demo today and discover how advanced automation eliminates bottlenecks, enabling enterprises to innovate at Michelin's scale. Learn why industry leaders rely on Delphix to deliver smarter, faster data throughout their entire lifecycle.

Request a data delivery demo today